In sub-Saharan Africa, cereal yields average about one tonne per hectare, and have barely risen in the past 30 years. In many countries, the situation is desperate. In Zambia, for example, 69% of smallholders can’t afford to buy mineral fertilizers, and around a third of the area planted with maize is abandoned each year. Declining soil fertility, coupled with the high price of mineral fertilizers, is largely to blame. But there is a low-cost remedy, and increasing numbers of farmers are benefiting from it. By planting green fertilizers – leguminous plants which draw nitrogen from the air to produce compounds which enrich the soil – farmers can restore fertility and increase yields.

Take, for example, Nelson Mkwaila, who farms a small plot of land near Blantyre, Malawi. “Ten years ago, I was lucky if I got one tonne of maize a hectare and I struggled to feed my family,” he recalls. “Now I get three times that much, thanks to these plants.” Mr Mkwaila is dwarfed not just by his maize, but by the Gliricidia bushes which grow between each row, acting as a fertilizer factory in his fields. Every year, before he sows his maize, he cuts back the Gliricidia; the leaves are incorporated into the soil and the woody stems provide fuel for the kitchen.

Sifting the evidence In Mr Mkwaila’s case, the fertilizer trees undoubtedly work. However, there has been considerable debate during recent years about the precise impact of woody and herbaceous legumes on soil fertility. “There’s been a lot of research on individual sites, but we needed to explain the variations in yield under different treatments,” explains Gudeta Sileshi, an agroecologist with the World Agroforestry Centre and senior author of Evidence for impact of green fertilizers on maize production in sub- Saharan Africa. “If farmers are to benefit from these technologies, it’s important that we understand the conditions under which these plants work best.” The meta-analysis conducted by Sileshi and his colleagues looked at the findings of 94 peer-reviewed studies. The increase in maize yields using green fertilizers was compared with the increase using mineral fertilizer, and with the yields of maize cropped continuously without fertilizer. “In broad terms, the use of green fertilizers increases yields,” explains Sileshi. The mean yield increase was highest at 2.3 tonnes per hectare for fully fertilized maize and ranged between 0.8 and 1.6 tonnes per hectare with green fertilizers. The meta-analysis found that the type of soil affects the degree to which green fertilizers increase yields, with the response being highest on nutrient-poor soils, and lowest on nutrient-rich soils. This means that green fertilizers offer the greatest benefits on land with low to medium potential, which is typically worked by poor farming families.

The use of green fertilizers significantly reduces the level of risk for farmers. In areas with low and erratic rainfall, green fertilizers reduce the likelihood of crop failure, with woody legumes making scarce water resources available to the maize crop. In areas which experience high rainfall and are prone to water-logging, green fertilizers improve the soil’s absorptive capacity and mop up some of the excess water. “Our analysis suggests there are also important synergistic effects when mineral fertilizers and legumes are used together,” says Sileshi. Maize yields increase by 25-30% when farmers use half the recommended dose of mineral fertilizers in tandem with green fertilizers. However, adding further quantities of fertilizer does little to improve yields further.

“This is a really substantial piece of work,” says Fergus Sinclair, global project leader for the World Agroforestry Centre’s research on increasing farm productivity. “It shows that fertilizer trees can lead to significant increases in yields under the right conditions.” The meta-analysis also opens up a new area of research. It is all very well showing that there is a mean increase in crop yields associated with the use of green fertilizers, but we now need to know what causes the variations around the mean. “Once we have the answers to that,” says Sinclair, “we will be able to refine the recommendations to farmers, and suggest which are the right legumes to use under which conditions.”